Introduction: Concurrent uterine lesions or an irregular endomyometrial junction can make accurate assessment of depth of myometrial invasion (DOI) and percentage of myometrial invasion (%MI) difficult, leading to patients being staged and or treated suboptimally. An alternative measurement, known as the tumor-free distance (TFD), which measures the distance between maximal myometrial invasion and the uterine serosa, has been proposed. Previous studies comparing the predictive abilities of DOI and TFD were underpowered and inconclusive. Our objective was to compare TFD, DOI, and %MI as predictors for lymph node involvement in surgically staged endometrial cancer patients.

Methods: Patients with endometrioid adenocarcinoma of the endometrium treated between January 1997 and December 2007 were included. Tumor-free distance, DOI, and %MI were evaluated along with other pathological variables to determine their predictive ability for nodal involvement. Depth of myometrial invasion was measured between the endomyometrial junction and the maximal myometrial invasion. Tumor-free distance was calculated by subtracting the DOI from myometrial thickness (MT). Percentage MI was derived by dividing DOI by MT and expressed as a percentage of MT invaded. These 3 variables were transformed to z scores, and their ability to predict nodal involvement was compared.

Results: A total of 338 patients were eligible for analysis. Mean (SD) MT was 18.7 (5.9) mm. Median DOI was 6 mm, and median TFD was 10.3 mm. On univariate analysis, all 3 variables showed significant associations with nodal involvement. On multivariate analysis, after adjusting for lymphovascular space invasion, cervical involvement, serosal/adnexal involvement, grade, %MI, and TFD, DOI retained its statistical significance along with lymphovascular space invasion and cervical involvement.

Conclusions: Depth of myometrial invasion predicts nodal involvement independently when compared with TFD.